dearwolf/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-05-20 02:24:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 21

chore(model gallery): add Deepseek-R1-Distill models (#4646)

Browse source 
* chore(model gallery): add Deepseek-R1-Distill-Llama-8b

Signed-off-by: Gianluca Boiano <morf3089@gmail.com>

* chore(model gallery): add Deepseek-R1-Distill-Qwen-1.5b

Signed-off-by: Gianluca Boiano <morf3089@gmail.com>

---------

Signed-off-by: Gianluca Boiano <morf3089@gmail.com>
This commit is contained in:
Gianluca Boiano 2025-01-21 10:37:05 +01:00 • committed by GitHub
parent 1a08948e63
commit b264a91b3f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 34 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

34
gallery/index.yaml
View file

@ -2696,6 +2696,23 @@
- filename: Qwentile2.5-32B-Instruct-Q4_K_M.gguf
sha256: e476d6e3c15c78fc3f986d7ae8fa35c16116843827f2e6243c05767cef2f3615
uri: huggingface://bartowski/Qwentile2.5-32B-Instruct-GGUF/Qwentile2.5-32B-Instruct-Q4_K_M.gguf
- !!merge <<: *qwen25
name: "deepseek-r1-distill-qwen-1.5b"
icon: "https://avatars.githubusercontent.com/u/148330874"
urls:
- https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5b
- https://huggingface.co/unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF
description: |
DeepSeek-R1 is our advanced first-generation reasoning model designed to enhance performance in reasoning tasks.
Building on the foundation laid by its predecessor, DeepSeek-R1-Zero, which was trained using large-scale reinforcement learning (RL) without supervised fine-tuning, DeepSeek-R1 addresses the challenges faced by R1-Zero, such as endless repetition, poor readability, and language mixing.
By incorporating cold-start data prior to the RL phase,DeepSeek-R1 significantly improves reasoning capabilities and achieves performance levels comparable to OpenAI-o1 across a variety of domains, including mathematics, coding, and complex reasoning tasks.
overrides:
parameters:
model: deepseek-r1-distill-qwen-1.5b-Q4_K_M.gguf
files:
- filename: deepseek-r1-distill-qwen-1.5b-Q4_K_M.gguf
sha256: c2c43b6018cf7700ce0ddee8807deb1a9a26758ef878232f3a142d16df81f0fe
uri: huggingface://unsloth/DeepSeek-R1-Distill-Qwen-1.5B-GGUF/DeepSeek-R1-Distill-Qwen-1.5B-Q4_K_M.gguf
- &archfunct
license: apache-2.0
tags:
@ -5219,6 +5236,23 @@
- filename: Dolphin3.0-Llama3.1-8B-Q4_K_M.gguf
sha256: 268390e07edd407ad93ea21a868b7ae995b5950e01cad0db9e1802ae5049d405
uri: huggingface://bartowski/Dolphin3.0-Llama3.1-8B-GGUF/Dolphin3.0-Llama3.1-8B-Q4_K_M.gguf
- !!merge <<: *llama31
name: "deepseek-r1-distill-llama-8b"
icon: "https://avatars.githubusercontent.com/u/148330874"
urls:
- https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B
- https://huggingface.co/unsloth/DeepSeek-R1-Distill-Llama-8B-GGUF
description: |
DeepSeek-R1 is our advanced first-generation reasoning model designed to enhance performance in reasoning tasks.
Building on the foundation laid by its predecessor, DeepSeek-R1-Zero, which was trained using large-scale reinforcement learning (RL) without supervised fine-tuning, DeepSeek-R1 addresses the challenges faced by R1-Zero, such as endless repetition, poor readability, and language mixing.
By incorporating cold-start data prior to the RL phase,DeepSeek-R1 significantly improves reasoning capabilities and achieves performance levels comparable to OpenAI-o1 across a variety of domains, including mathematics, coding, and complex reasoning tasks.
overrides:
parameters:
model: deepseek-r1-distill-llama-8b-Q4_K_M.gguf
files:
- filename: deepseek-r1-distill-llama-8b-Q4_K_M.gguf
sha256: f8eba201522ab44b79bc54166126bfaf836111ff4cbf2d13c59c3b57da10573b
uri: huggingface://unsloth/DeepSeek-R1-Distill-Llama-8B-GGUF/DeepSeek-R1-Distill-Llama-8B-Q4_K_M.gguf
- &deepseek ## Deepseek
url: "github:mudler/LocalAI/gallery/deepseek.yaml@master"
name: "deepseek-coder-v2-lite-instruct"